Researchers introduce PolyWorkBench, a benchmark designed to evaluate large language model agents on multilingual long-horizon workplace workflows. This addresses the gap in existing benchmarks that typically assume monolingual settings, whereas real-world applications often involve mixed-language inputs and outputs.

  • The benchmark comprises 67 tasks across five domains: commerce, knowledge work, legal analysis, localization, and manufacturing.
  • Agents are required to process heterogeneous multilingual inputs, perform iterative reasoning, invoke external tools, and produce structured outputs.
  • Evaluation utilizes a hybrid framework combining structural grading, executable verification, and LLM-based semantic assessment.
  • Empirical results indicate that state-of-the-art LLM agents suffer significant performance degradation in multilingual settings compared to monolingual ones.

The analysis suggests that multilinguality introduces compounding effects across reasoning and execution steps, highlighting the need for jointly modeling language variation and procedural decision-making.